Embed Size (px)
Citation preview
Network Slicing Use Case
Requirements April 2018
3
Executive SummaryCreating smart 5G networks
Network Slicing is set to be a prominent feature
of 5G to allow connectivity and data processing
tailored to specific customers’ requirements.
Mobile communications provided by smart
networks will enhance the efficiency and
productivity of business processes and will
open up opportunities for operators to address
the Business to Business segment more
effectively.
The Business Opportunity size
In the 5G era, different industry verticals are
seeking to leverage the power of technology to
boost productivity across swathes of the
economy. Network Slicing builds on this
expectation, and together with the promise of
Massive IoT and ultra-reliable/low latency
services, will support the digital transformation
and mobilisation of industry vertical customers.
The GSMA estimates that in combination with
other enablers and capabilities, Network
Slicing will permit operators to address a
revenue opportunity worth $300bn by 2025.
Understanding the customers’ needs
The possibility of tailoring mobile network
properties to the needs of the business through
the configuration of a large set of parameters
offers unsurpassed flexibility. However, with
such a diverse range of possible requirements
from verticals, operators will need to manage
risks from excessive complexity in the service
offering and cumbersome management,
leading to higher costs.
In this paper, the GSMA sets out to understand
the service requirements expressed by
business customers from different vertical
industries in several key sectors including
energy, IoT, automotive, manufacturing and
many more.
Requirements for each use case were
analysed, quantified where possible and
categorised into performance, functional and
operational requirements.
Defining a common starting point
This paper proposes the adoption of a “generic
slice template” that the industry can use as
universal description. A generic slice template
provides a universal description of a network
slice type that can be used by infrastructure
vendors, mobile network operators and slice
buyers. When populated with values for all or a
subset of the attributes, the generic slice
descriptor can serve many purposes:
Infrastructure vendors, can use the
descriptor to define the features of their
products
The slice buyer can use the descriptor as a
reference for SLA/contractual agreements
with the operator
Operators can exchange slice descriptors
with their roaming partners facilitating the
support of service continuity when moving
between networks.
The generic slice template will also serve as a
baseline for the definition of a set of
standardised service/slice types.
This study has also been used to identify the
use cases that can be served by the same
network slice type as well as use cases that
are likely to require simultaneous support,
something referred to as a “network slice
bundle”. A network slice bundle describes the
family of network slice types required to serve
a group of use cases, for example a network
slice bundle for a vehicle will include slices
supporting; a high-reliability telemetry service,
a high bandwidth infotainment service, a low
latency “vehicle to network to X” service, a
super low latency vehicle to vehicle, etc.
4
Acknowledgements Special thanks to the following GSMA Network Slicing Taskforce members for their contribution
and review of this document:
AT&T Mobility
BlackBerry Limited
China Mobile Limited
China Telecommunications Corporation
China Unicom
Cisco Systems
Deutsche Telekom AG
EE Limited
Emirates Telecommunications Corporation (ETISALAT)
Ericsson
Gemalto NV
OPPO
Hong Kong Telecommunications (HKT) Limited
Huawei Technologies Co Ltd
Hutchison 3G UK Limited
Intel Corporation
Jibe Mobile Inc.
KDDI Corporation
KT Corporation
Kuwait Telecom Company (K.S.C.)
MediaTek Inc.
Metaswitch Networks
Nokia
NTT DOCOMO Inc.
Orange
Qualcomm Incorporated
Radiomóvil Dipsa
S.A. de C.V.
Samsung Electronics Co Ltd
SK Telecom Co. Ltd.
Sprint Corporation
Starhome Mach
Syniverse Technologies Inc.
Telecom Italia SpA
Telefónica S.A.
Telenor Group
Telia Finland Oyj
United States Cellular Corporation
Verizon Wireless
Vodafone Roaming Services
ZTE Corporation
5
ContentsExecutive Summary ................................ 3
Acknowledgements ................................ 4
Contents .................................................. 5
Introduction ............................................. 7
High-level view on Network Slicing .................7
Abbreviations .....................................................8
References ...................................................... 10
Terminology .................................................... 11
Network Slicing: exploring the business
case ........................................................ 13
Sizing the Network Slicing opportunity ........ 13
Network Slicing is part of the investment case
for 5G ............................................................... 14
Go-to-market Strategy .................................... 14
Avoiding the commercial shortcomings of
previous QoS type solutions ......................... 15
Standardisation Landscape ................. 17
Vertical Industry Organisation ...................... 17
Telecom Industry Organisation ..................... 17
Standards Developing Organisation ............ 18
Telecom Industry Organisation ........... 21
AR/VR ..................................................... 21
Overview .......................................................... 21
Use cases ........................................................ 21
Automotive ............................................ 23
Overview .......................................................... 23
Use Cases ........................................................ 23
Summary of requirements ............................. 25
Energy .................................................... 25
Overview .......................................................... 25
Use Cases ........................................................ 25
Summary of requirement ............................... 27
Healthcare ............................................. 27
Overview .......................................................... 27
Use cases ........................................................ 27
Summary of requirements ............................. 27
Industry 4.0, Manufacturing ................. 28
Overview .......................................................... 28
Use cases ........................................................ 28
Summary of requirements ............................. 29
Internet of Things for Low Power Wide
Area Applications ................................. 29
Overview .......................................................... 29
Use cases ........................................................ 29
Summary of requirements ............................. 30
Public Safety .......................................... 31
Overview .......................................................... 31
Use Cases ........................................................ 31
Summary of Requirements ............................ 32
Smart Cities ........................................... 32
Overview .......................................................... 32
Use cases ........................................................ 32
Summary of requirements ............................. 33
Industry recommendations and
reshaping the Network Slicing concept
35
Requirements analysis ................................... 35
Operational requirements .............................. 39
Generic Network Slicing template ....... 41
Cross-Standardisation Collaboration
Recommendations ................................ 42
Coordination from System Architecture
Perspective ...................................................... 42
Coordination from Management Perspective
.......................................................................... 42
Coordination from Security Perspective ...... 43
Coordination between SDOs and Open
Source .............................................................. 43
Other recommendations ....................... 44
Network Slicing Concept Evolution ..... 44
Operational isolation vs. network level
isolation ........................................................... 44
Not a network slice for each vertical customer
.......................................................................... 45
Number of slices supported by the 5G system
.......................................................................... 45
Network slice deployment ............................. 45
Regulatory aspects ............................... 48
Net Neutrality ................................................... 48
Quality of Service............................................ 48
Cross-Border Data Transfers ........................ 49
Illegal Content ................................................. 50
Summary .......................................................... 51
Next steps .............................................. 53
Annex A Document Management ....... 54
6
1
Introduction
7
Introduction
5G networks, in combination with Network
Slicing, permit business customers to enjoy
connectivity and data processing tailored to
their specific business requirements. It is
expected that providing tailored services to
business customers has significant commercial
potential.
The main purpose of this document is to
provide a comprehensive overview about the
service requirements expressed by business
customers from different vertical industries -
consumers are purposefully not addressed.
These requirements have been collected
through discussions and interviews as well as
analysis of available white papers. In addition,
these requirements were analysed in detail and
industry recommendations are deviated
thereof. The paper also provides an overview
of the standardisation landscape of Network
Slicing, business model considerations and
regulatory needs. Security, privacy and liability
have also been covered in this paper. Finally,
the generic slice template concept is briefly
introduced which contains all the potential
attributes a network slice could have. This
gives us a baseline for all network slices
offered to customers by specifying the values
that each parameter would take for a given
slice instance.
This document is the successor of previous
documents [1], [2] which are written from the
customer perspective and which define what
Network Slicing is and what Network Slicing
could offer to the customers.
The document is structured as follows:
Chapter 1 introduces the paper and
provides a brief summary of the Network
Slicing definition coming from the previous
papers and the terminology used
throughout the document.
Chapter 2 explores the business cases for
Network Slicing, which could potentially
change the way network operators do their
business and to enable new business
models.
Chapter 3 gives a snapshot of the
standardisation landscape with respect to
Network Slicing and covers vertical industry
organisations, telecom industry
organisations as well standards developing
organisations.
Chapter 4 summarises the results of the
study the GSMA conducted on the service
requirements analysis for different vertical
industries.
Chapter 5 represents the core part of this
document, which aims to analyse the
service requirements and provides
recommendations towards the industry. It
also discusses the idea of the generic slice
template.
Chapter 6 provides an overview about the
policy aspects around Network Slicing such
as net neutrality.
Chapter 7 concludes the document and
provides details about the planned next
steps.
High-level view on Network Slicing
From a mobile operator’s point of view, a
network slice is an independent end-to-end
logical network that runs on a shared physical
infrastructure, capable of providing an agreed
service quality. The technology enabling
Network Slicing is transparent to business
customers for whom 5G networks, in
combination with Network Slicing, allow
connectivity and data processing tailored to the
specific business requirements. The
customisable network capabilities include data
speed, quality, latency, reliability, security, and
services. These capabilities are always
8
provided based on a Service Level Agreement
(SLA) between the mobile operator and the
business customer.
A network slice could span across multiple
parts of the network (e.g. access network, core
network and transport network) and could be
deployed across multiple operators. A network
slice comprises of dedicated and/or shared
resources, e.g. in terms of processing power,
storage, and bandwidth and has isolation from
the other network slices.
It is anticipated that mobile network operators
could deploy a single network slice type that
satisfies the needs of multiple verticals, as well
as multiple network slices of different types that
are packaged as a single product targeted
towards business customers (a business
bundle) who have multiple and diverse
requirements (for example a vehicle may need
simultaneously a high bandwidth slice for
infotainment and an ultra-reliable slice for
telemetry, assisted driving).
Abbreviations
Term Description
3GPP Third Generation Partnership
Project
3GPP RAN Third Generation Partnership
Project Radio Access Network
3GPP SA Third Generation Partnership
Project Service & Systems
Aspects
4G 4th Generation Mobile Network
5G 5th Generation Mobile Network
5GAA 5G Automotive Association
AMF Access and Mobility Management
Function
APEC Asia-Pacific Economic
Cooperation
API Application Programming
Interface
AR Augmented Reality
AN Access Network
BBF Broadband Forum
C2C Control-to-control
CN Core Network
CV2X Cellular vehicle-to-everything
D2D Device to device
DiffServ Differentiated services
E2E End-to-End
EPS Evolved Packet System
ETSI European Telecommunications
Standards Institute
EU European Union
FMC Fixed mobile convergence
eMBB Enhanced Mobile Broad Band
GDP Gross Domestic Product
GPS Global Positioning System
GST Generic Slice Template
HMI Human-machine interface
ICT Information and Communication
Technologies
IEEE The Institute of Electrical and
Electronics Engineers
IETF Internet Engineering Task Force
IoT Internet of Things
IntServ Integrated Services
IIC Industrial Internet Consortium
IP Internet Protocol
IPR Intellectual Property Rights
ISG Industry Specification Group
ISP Internet Service Provider
ITU International Telecommunication
Union
ITU-T ITU Telecommunication
Standardisation Sector
KPI Key Performance Indicator
L3VPN Layer 3 VPN
9
LPWA Low Power Wide Area
LTE Long Term Evolution
MANO Management and Orchestration
MBB Mobile Broadband
MCPTT Mission Critical Push to Talk
MDM Multiple Dedicated Networks
MEC Multi-access Edge Computing
MEF Metro Ethernet Forum
mMTC Machine Type Communications
MTC Machine Type Communication
NB-IoT Narrow Band IoT
NC New Core
NFV Network Function Virtualisation
NF Network Function
NGFI Next Generation Fronthaul
Interfaces
NGMN Next Generation Mobile Network
NR New Radio
NSI Network Slice Instance
OEM Original Equipment Manufacturer
OSM Open Source Management and
Orchestration (MANO)
ONAP Open Network Automation
Platform
OTT Over The Top
PRD Permanent Reference Document
PPDR Public Protection and Disaster
Recovery
QCI Quality of Service Class Identifier
QoS Quality of Service
RTT Round-trip Time
SDN Software Defined Networks
SDO Standards Developing
Organisation
SLA Service Level Agreement
SLR Service Level Reporting
SR Segment Routing
SST Service Slice type
TN Transport Network
TETRA Terrestrial Trunked Radio
UE User Equipment
UPF User Plane Function
URLLC Ultra Reliable Low Latency
Communications
V2X Vehicle to X (e.g. Vehicle,
Infrastructure, Pedestrians)
VoLTE Voice over LTE
VPN Virtual Private Network
VR Virtual Reality
WG Working Group
ZVEI Zentralverband Elektrotechnik
und Elektronikindustrie -
Germany's Electrical Industry
10
References
Ref Title
[1] Smart 5G networks: enabled by Network Slicing and tailored to customers’ needs
[2] An Introduction to 5G Network Slicing
[3] http://5gaa.org/
[4] https://www.zvei.org/en/
[5] http://www.iiconsortium.org/
[6] https://www.tmforum.org/
[7] https://www.broadband-forum.org/
[8] http://www.etsi.org/technologies-clusters/technologies/nfv
[9] https://portal.etsi.org/tb.aspx?tbid=862&SubTB=862
[10] 5GPPP Project 5GCAR, “Deliverable D2.1: 5GCAR Scenarios, Use Cases, Requirements
and KPIs”.
[11] 5G-PPP White Paper on Automotive Vertical Sector, “5G Automotive Vision”, Oct. 2015
[12] 3GPP TR 22.886 V15.1.0, “Study on enhancement of 3GPP Support for 5G V2X Services”
[13] 3GPP TS 22.186 V15.2.0, “Enhancement of 3GPP support for V2X scenarios”
[14] 3GPP TS 22.185 V14.3.0, “Service requirements for V2X services”
[15] 5GPPP White paper, “5G and Energy”, Sep 2015
[16] 3GPP TS 22.261 V16.1.0, “Service requirements for the 5G system”
[17] Perspectives on Vertical Industries and Implications for 5G”, 06/2016.
https://www.ngmn.org/fileadmin/user_upload/160610_NGMN_Perspectives_on_Vertical_In
dustries_and_Implications_for_5G_v1_0.pdf
[18] 3GPP TS 22.804, “Study on Communication for Automation in Vertical Domains”
[19] What is the Internet of Things
https://www.gsma.com/iot/wp-content/uploads/2016/09/What-is-the-Internet-of-Things.pdf
[20] 3GPP Low Power Wide Area Technologies
https://www.gsma.com/iot/wp-content/uploads/2016/10/3GPP-Low-Power-Wide-Area-
Technologies-GSMA-White-Paper.pdf
[21] ETSI Multi-access Edge Computing (MEC)
http://www.etsi.org/technologies-clusters/technologies/multi-access-edge-computing
[22] ONAP https://www.onap.org/
[23] OSM: https://osm.etsi.org/
[24] ETSI ISG “MEC support for network slicing
https://portal.etsi.org/webapp/WorkProgram/Report_WorkItem.asp?WKI_ID=53580
[25] Zero-touch Network and Service Management – Introductory White Paper, December 7,
2017
https://portal.etsi.org/TBSiteMap/ZSM/OperatorWhitePaper
[26] NGMN Alliance, “5G End-to-End Architecture Framework”, 10/2017.
[27] Huawei iLab, “Cloud VR–oriented Bearer Network White Paper”, 2017.11.02
11
[28] 3GPP TS 33.501, 0.7.0, “Security architecture and procedures for 5G System”
[29] http://www.ieee802.org/1/pages/802.1as.html
[30] https://standards.ieee.org/findstds/standard/1588-2008.html
[31] 5G Network Slicing for Vertical Industries
https://gsacom.com/paper/5g-network-slicing-vertical-industries/
[32] China Telecom, China’s State Grid, and Huawei, “5G Network Slicing Enabling the Smart
Grid”, January 22, 2018
Terminology
Network Slice: A network slice is a logical
network that provides specific network
capabilities and network characteristics in
order to serve a defined business purpose
of a customer. Network Slicing allows
multiple virtual networks to be created on
top of a common shared physical
infrastructure. A network slice consists of
different subnets, example: Radio Access
Network (RAN) subnet, Core Network (CN)
subnet, Transport network subnet.
Network Slicing provider: Typically a
telecommunication service provider, is the
owner or tenant of the network
infrastructures from which network slices
are created. The Network Slicing provider
takes the responsibilities of managing and
orchestrating corresponding resources that
the Network Slicing consists of.
Business customer: A business customer
tenants the network slice, e.g. customers
from vertical industries. For instance, a
business customer could be an enterprise
or specialised industry customer (often
referred to as “verticals”).
12
2 Network Slicing:
exploring the business
case
13
Network Slicing: exploring the business case
Sizing the Network Slicing opportunity
Network Slicing is integral to unlocking the
enterprise opportunity ($300bn by 2025) for
the 5G era. Already, the industry expects the
enterprise segment to be the most important
source of incremental revenue opportunity in
the 5G (Figure 1). Network Slicing is one of
the 5G era tools and enablers that will
support operators to unlock the enterprise
opportunity.
In the 5G era, different industry verticals are
seeking to leverage the power of technology
to boost productivity across swathes of the
economy. Network Slicing builds on this
expectation, and together with the promise of
Massive IoT and ultra-reliable/low latency
services, will support the transformation of
vertical industries.
To unlock this opportunity, Network Slicing
will enable operators to create pre-defined,
differing levels of services to different
enterprise verticals, enabling them to
customise their own operations.
However, the opportunity could become even
bigger. Automation and the ability to quickly
create slices could pave the way for
operators to dynamically package and
repackage network capabilities for different
customers. This is the end goal of Network
Slicing.
Figure 1: New incremental revenue opportunities in 5G to come from the enterprise segment
Figure 2: The role of collaboration in the investment case for Network Slicing
14
Network Slicing is part of the investment
case for 5G
There are three considerations about the
investment case for Network Slicing,
especially given the potential complexity and
cost of the required automation.
Firstly, ongoing industry activities to
standardise Network Slicing should gear
towards minimising the complexity of the
technical solution so that adoption can be
made relatively easy.
Secondly, the GSMA is working with its
members to streamline the commercial
deployment scenario for Network Slicing in
order to drive economies of scale and reduce
unitary cost of deployment.
Thirdly, operators and their vendors are
working to make the cost of deploying
Network Slicing marginal to the broader
investment case for 5G. This is an important
consideration because building a new
investment case to retrofit an already
deployed network is difficult and slow. That is
the lesson learned from the past where
VoLTE launches were decoupled from LTE
launches (compared to 2G/3G voice
services).
Go-to-market Strategy
There are three stages to the Network Slicing
go-to-market strategy. These stages deliver
value across the Value Chain, and may
happen in parallel depending on local market
conditions:
Deploy Network Slicing for internal use: This
will prove the validity of Network Slicing by
using the solution to serve internal customers
within an operator or the operator’s sister
Figure 3: Stages of the Network Slicing go-to-market strategy
Figure 4: Factors that could impact commercial adoption of Network Slicing
15
companies. This option offers a low risk
opportunity to experiment and to validate the
proposition in order to refine it ahead of
rollout to commercial customers.
Upsell Network Slicing capabilities to existing
enterprise customers: This will prove value to
existing enterprise customers and based on
the typical buying behaviour of business
customers, upselling Network Slicing
capabilities to these customers ought to be
an easier opportunity than targeting new
customers. These customers can then
become proof points and advocates for the
new capabilities.
Sell to new enterprise customers: When
commercially ready, slicing will be made
more broadly available to enterprise
customers who often require a proven
solution and seek market validation before
they buy.
Avoiding the commercial shortcomings of
previous QoS type solutions
The telecoms industry has had a history of
creating technical solutions that can offer
differentiated service levels to different
customers e.g. DiffServ, IntServ, QCI etc. Yet
despite fully developed technical solutions,
there is no evidence of widespread
commercial use of any of these solutions in
the industry.
If Network Slicing is to avoid the same
shortcomings, the industry needs to move
fast to overcome the technical, commercial
and regulatory factors that could impede
progress.
16
3 Standardisation
landscape
17
Standardisation Landscape
Network Slicing is a concept for running
multiple logical customised networks on a
shared common infrastructure complying with
agreed Service Level Agreements (SLAs) for
different vertical industry customers (or
tenants) and requested functionalities. To
achieve this goal, Network Slicing needs to be
designed from an End-to-End (E2E)
perspective, spanning over different technical
domains (e.g., access network, core network,
transport network and network management
system) as well as administrative domains
(e.g., pertaining to different mobile network
operators). Figure 5 illustrates a snapshot of
various groups and organisations relevant to
Network Slicing development.
Even though many organisations are tackling
Network Slicing issues as discussed above,
they also notice that Network Slicing issues
cannot be resolved in a single entity.
Therefore, they should be liaised between
them. The relationship among different
industry platforms and SDOs is shown in
Figure 6.
In the following sections more details about
the vertical industry and telecom industry
organisations as well as the standards
developing organisations will be provided.
Vertical Industry Organisation
Network Slicing aims to support various
vertical industries in the 5G era, hence the
direct requirements from vertical industries
are essential for Network Slicing designs. 5G-
awareness has been developed among
different vertical industries. For instance, 5G
Automotive Association (5GAA), created in
September 2016, aims at a global, cross-
industry Organisation of companies from the
automotive, technology, and
telecommunications industries (ICT), working
together to develop end-to-end solutions for
future mobility and transportation services [3].
In Nov 2017, the first Network Slicing
workstream was established in 5GAA WG5 to
understand the business aspects of Network
Slicing in automotive industry. Manufacturing
industry organisations like Zentralverband
Elektrotechnik und Elektronikindustrie (ZVEI)
[4] and Industrial Internet Consortium (IIC) [5]
also started to engage with 5G for next
generation smart manufacturing solutions.
Telecom Industry Organisation
Telecom Industry Organisations like the
GSMA and NGMN (Next Generation Mobile
Networks) describe the business drivers,
concepts, and high-level requirements of E2E
Network Slicing from the operator’s point of
view. The GSMA has initiated the Network
Figure 5: Network Slicing Relevant Industry Groups and SDOs Landscape
18
Slicing Taskforce (NEST) project to
harmonise slicing definition, identify slice
types with distinct characteristics and
consolidate parameter and functionality
requirements. The NEST aims at generating
a Permanent Reference Document (PRD) to
guide future Network Slicing standards.
The NGMN Alliance is developing,
consolidating and communicating
requirements to ensure that customer needs
and expectations on mobile services are
fulfilled. The Alliance actively drives global
alignment and convergence of technology
standards and industry initiatives with the
objective to avoid fragmentation and to
guarantee industry scale.
TM Forum ZOOM project [6] has started a
workstream to analyse Network Slicing
business models and business scenarios of
high interest to service providers, vertical
industries, and other potential Network
Slicing consumers. A number of user stories
have been generated, and respective
requirements have been derived and mapped
to TM Forum Assets.
Standards Developing Organisation
Various technologies and innovations from
different technical domains have substantially
contributed to the Network Slicing progress in
different Standards Developing Organisations
(SDO). Currently, technical specifications of
those domains are defined in corresponding
SDOs, namely, Radio Access Network (RAN)
and Core Network (CN) by 3GPP, Transport
Network (TN) by BBF and IETF, etc. ITU-
T(GSTR-TN5G), IEEE(NGFI 1914), MEF and
other SDOs are working on this topic as well.
The major SDOs will be introduced in the
following sections.
3GPP
3GPP could be considered as the
forefront ambassador for Network Slicing.
There are many Working Groups in 3GPP
related with Network Slicing. In 3GPP
SA1, Network Slicing related use cases
and requirements are defined. In 3GPP
RAN1/2/3, the RAN slicing-awareness
features are discussed. In 3GPP SA2, the
fundamental system architecture choice to
support Network Slicing is specified. For
Figure 6: SDO relationship figure
19
instance, the fundamental network
functions and procedures to support slice
selection/access, session management
within network slices, as well as the
terminal behaviour, e.g. a UE could
support maximum eight slices in parallel,
and a common core network function is
defined, i.e. AMF, for one UE to be
capable to use all slices.
The 3GPP SA5 Working Group are
responsible for completing the creation
and management of slices in the 3GPP
realm and for driving the coordinating with
other relevant SDOs to generate complete
E2E network slices. SA3 Working Group
[28] are responsible for security
capabilities of E2E Network Slicing that
require triggering and coordination with
ETSI ISG NFV on isolation of Network
Slices.
BBF
BBF’s [7] main scope in Network Slicing is
to clarify the requirements for 5G bearer
networks and defines the related TN
slicing management architecture. The
BBF should establish formal cooperation
with 3GPP to facilitate the transmission
requirements from 3GPP and coordinate
the interface requirements between the
3GPP slicing management system and
Bearer slicing management system,
including corresponding slicing creation
and management processes. The
technical definition for a specific interface
is not within the scope of the BBF
standardisation, but they can recommend
the potential options.
IETF
To support Network Slicing, the IP router
should be enhanced, which will extend
existing protocols, such as Segment
Routing (SR) and L3VPN. Furthermore,
potentially DetNet is a candidate
technology for transporting traffic of
URLLC slice type. Extension or new
mechanism may be standardised to meet
such requirement.
Another work in IETF is primary to define the
interface between the 3GPP slicing
management system and TN slicing
management system. For the IP transport
network, this interface is the northbound
interface of TN slicing management system
defined in BBF.
ETSI
ETSI ISG NFV
The ETSI NFV ISG (Network Function
Virtualisation) [8] is responsible for
providing technical solutions for
resources such as computing and
storage for Network Slicing. They
consider NFV is a key enabler for the
5G infrastructure. However, further
work is still required to understand
how NFV technology supports
Network Slicing, analyses the impacts
to the current NFV architecture and
standardises the NFV slicing resource
layer. Including security and reliability-
related features such as virtualisation
level security isolation scheme,
function orchestration of virtual border
gateway (such as firewall),
deployment and configuration
strategy.
ETSI ISG ZSM
In Dec 2017, ETSI created the Zero
touch network and Service
Management Industry Specification
Group (ZSM ISG) [9]. This initiative
aims to resolve the 5G E2E Network
Slicing management issue. Their
vision is to enable full automation in
terms of delivery, deployment,
configuration, assurance, and
optimisation of network services.
20
4 Vertical Requirements
21
Telecom Industry Organisation
This chapter briefly summarises different
vertical industries, their most important use
cases and their most important service
requirements. The intention is not to provide
very detailed information, but instead give an
idea about the biggest challenges for each
vertical industry. References to more detailed
information are provided where available.
It should be noted that the list of verticals is
not complete but can be considered as a
snapshot of the current status. In addition,
the status of the different verticals is different
while some have very concrete ideas about
use cases and requirements with concrete
values associated to them some have vague
ideas with initial indications. More
discussions and analysis are required in the
future to complement the list of verticals as
well as details about the requirements and
use cases.
AR/VR
It should be noted that AR/VR is not a vertical
industry in itself. Instead AR/VR use cases
can be found in almost all of the vertical
industries. However, these use cases are so
important that they are discussed in a
separate section.
Overview
Augmented Reality (AR) is a technique where
the real world view is augmented, or assisted,
by a computer generated views, this can be
in single or multi-sensory modes including,
auditory, visual, and haptic.
Virtual Reality (VR) is the technology to
construct a virtual environment, which may
be based on the real environment within
which people could have real-time
interaction. There are a number of key
technologies used together to enable VR, i.e.
360 degree panorama video, Freeview-point,
computer graphics, light field, etc. Many
applications are derived from VR, for
instance, VR gaming, VR broadcasting, VR
simulated environment for education,
healthcare, military training, etc. Moreover,
the mutuality of the Cloud capability (e.g.
rendering in cloud, edge computing) as well
as pervasive network infrastructure
deployment will bring VR applications to a
new dimension.
Use cases
Strong-Interactive VR: Audio-visual
interaction
Audio-visual interaction is characterised by a
human being interacting with the environment
or people, or controlling an UE, and relying
on audio-visual feedback. In order to achieve
the VR environments with low motion-to-
photon requirements, the 5G network slice
should have the capability of motion-to-
photon latency in the range of 7-15ms while
maintaining the 250Mbps user data rate and
motion-to-sound delay less than 20ms.
During voice conversions, there exists some
interactive tasks, 5G Network Slicing is
required to have the capability to support
100ms one-way mouth-to-ear low-delay
speech coding for interactive conversational
services.
Furthermore, audio components and video
components are usually handled separately,
so in order to avoid having a negative impact
on the user experience, the communication
system will also have to cater for the VR
audio-video synchronisation. The audio-video
synchronisation thresholds is 5-125ms for
audio delayed and 5-45ms for audio
advanced.
22
Strong-Interactive VR: Low-delay speech
and video coding
Current speech codecs have an inherent
coding delay of 20-40ms. In a 4G network,
such coding delay is not a problem during a
phone call because even a 400ms one-way
delay between speakers does not seriously
impair an interactive discussion. However,
when the speech voice is used in a highly
interactive environment, e.g., a multiplayer
game or a virtual reality meeting, the
requirements on the speech coding delay
become tougher to meet, and current coding
delays are too high. To support interactivity,
the one-way delay for speech should be
10ms (or lower). These scenarios have a
critical requirement on transfer bandwidth
and delay to guarantee good user experience
compared to current video service.
In order to fulfil the performance
requirements of low-delay speech and video
coding, the design of the network slice should
consider different aspect of video codec, e.g.
frame rates, resolution and bandwidth. The
higher the frame rate (fps - frames per
second) the better the video quality, virtual
reality may require the capability of displaying
content at frame rates of 120 fps or more.
Strong-Interactive AR: Use Cases
Assisting with complex tasks: Use of AR to
overlay instructions can speed up fault finding
and repairs by utilising heads up display with
connected glasses.
Connecting remote workers: Allowing remote
experts to see exactly what the worker sees
and direct them to complete tasks. Enabling a
deeper level of assistance and guidance to
be conveyed.
Accelerated Learning: AR training packages
can facilitate vastly more effective training.
Supervisors are able to mentor and assess
capability, resulting in higher-quality work
with less mistakes.
Summary of requirements
In order to understand VR applications’
requirements for Network Slicing, VR
applications could be classified into two
categories according to how an end user
interacts with the virtual environment: weak-
interactive VR (wi-VR) and strong-interactive
VR (si-VR). According to the technology
maturity, VR use cases may experience three
different stages:
entry-level VR (now-2 years)
advanced VR (3~5 years)
ultimate VR (6-10years).
Weak-interactive VR
For wi-VR, end users do not have direct
interaction with the virtual environment, the
VR content is pre-planned and end users
may be possible to select the observation
point and location. Therefore, end users’
experience for wi-VR is passive.
Strong-Interactive VR
With the progression of VR technology, end
users could interact with the virtual
environment via specific user interface
(software or hardware) and the virtual
environment will provide real-time response,
Entry-Level VR
(FOV,8K 2D/3D)
Advanced VR
(FOV,12K 3D)
Ultimate VR
(FOV,24K 3D)
Data rate 40Mbps（2D）
63Mbps（3D）
340Mbps
FOV：2.34Gbps
Typical RTT 30ms（2D）
20ms（3D）
20ms
10ms
Packet loss 2.40E-5 1.00E-6 1.00E-6
Table 1: Network performance requirements from wi-VR[27]
23
which brings the interactive and immersive
user experience. The VR content may be
different according to user’s input; hence, the
content may be not planned in advance.
Automotive
Overview
Cellular vehicle-to-everything (CV2X) is
considered as one of the most prominent use
cases for 5G, and the automotive industry
has been one of the first verticals that
engaged with the communications technology
industry. CV2X aims to enable
communication amongst vehicles (referred to
Vehicle to Vehicle or V2V) as well as
communication between vehicles and an
infrastructure. Not only CV2X will create new
business opportunities for OEMs, e.g.
providing in-car infotainment service, but will
also increases road safety. To support
performance critical type of services (e.g.
autonomous/semi-autonomous/assisted
driving), CV2X communication also includes
other connected-vehicle technology, e.g. in-
car sensors, cameras and radar systems.
Providing support for V2X services using
operator’s 5G public network, not only will
significantly reduce deployment cost
compared to running V2X services on a
dedicated network, but will also offer a better
coverage. Moreover, CV2X can benefit from
support of critical performance (e.g. ultra-
reliable and low latency communication) and
customised network services (e.g. network
function could be tailored according to
customers’ requirements) that are included
by design in 5G.
However, for a successful deployment of
CV2X, 5G operators need to have a full
understanding of the use cases and
requirements of the automotive industry.
Use Cases
5G CV2X use cases have been well studied
by many research programs (e.g. 5GPPP
[10][11]), standardisation (e.g. 3GPP [12][13])
and industry oraganisations (e.g. 5GAA [3]).
A number of representative use case
families/classes are presented in the
following sub-sections.
Infotainment
This type of use cases requires direct
data exchange between vehicle and
application servers via mobile systems.
Such services normally focus on providing
a more pleasant driving experience both
for driver and passengers so they are not
safety-critical and can be delivered using
mobile broadband (MBB) connectivity.
Examples include: music, movies, live TV
streaming, audio/video conference
streaming (office-in-car), online gaming,
web browsing.
Telematics
This type of use case also requires direct
data exchange between vehicle and
application servers via a mobile system,
Entry-Level VR
(FOV,8K 2D/3D)
Advanced VR
(FOV,12K 3D)
Data rate 120Mbps（2D）
200Mbps（3D）
1.40Gbps
Data rate
Typical RTT 10ms 5ms
Typical RTT
Packet loss 1.00E-6 1.00E-6 Packet loss
Table 2: Network performance requirements from si-VR[27]
24
which are normally provided by
automotive manufacturer (or their
authorised third-party service provider).
Being different from the above category, it
provides services to assist the driving
experience. Example use cases are
navigation provision, remote health
monitoring of the vehicle, precise position
provisioning, parking slot discovery,
automated parking, etc. An automotive
manufacturer could also use the
connectivity to schedule a control module
firmware and software update over mobile
system for selected range/type of
vehicles.
Road Safety and Efficiency
Basic Safety Services
Basic safety services support the
driver with additional information to
improve road safety and efficiency.
Road Warning
These services provide information to
the driver about imminent dangers
such as red light violation, hazard
warning, forward collision warning,
intersection collision warning, traffic
jam warning, etc. Such information
could help the drivers to take remedial
actions (e.g. lane changing,
deceleration). Being different from the
driving experience assistance
mentioned in Telematics, these use
cases normally are triggered by a
specific event (e.g. based on real time
road situation detection) and the
actions are taken by driver, hence it is
not strictly delay-sensitive but it is
preferred to be delivered as fast as
possible.
Information (Sensor data) Sharing
Perception of surrounding
environment for driving condition
analysis is a very crucial aspect to
improve driving safety. Real-time
information could be exchanged
among vehicles, e.g. on-vehicle
sensor or information captured by the
vehicles like traffic information. Such
information could be used for
collective perception of environment to
avoid potential dangers. One typical
use case is “See-Through”, which
refers to camera and/or radar data
sharing to improve/extend driver’s
visibility. For instance, by receiving a
video stream captured by the front
truck can help the driver behind to
make an overtaking decision.
Information sharing is also one of the
essential factors to enable
autonomous driving services
mentioned below.
Advanced Driving Service
Advanced driving services enable
semi-automated or fully automated
driving.
Cooperative driving
Information (on-vehicle sensor data
and driving actions like braking and
accelerating) can be exchanged
among vehicles for cooperative
collision avoidance. Another example
use case is cooperative lane merging,
where vehicles exchange information
on their intended trajectories and
perform automated lane changing
manoeuvres to avoid collisions and to
improve traffic flow.
Platooning
This use case class describes
operating a group of vehicles in a
closely linked manner so that the
vehicles move like a train with virtual
strings attached between vehicles. To
keep the vehicles as close as possible
with safety assurance, the vehicles
need to share status information (such
as speed, heading) as well as their
driving intentions (such as braking,
acceleration). By doing this, the overall
fuel consumption could be lowered
25
down and the amount of required
drivers can also be reduced.
Tele-operation
This use case class describes the
scenarios where a driver could directly
control an autonomous-capable
vehicle from a remote location such as
a control centre in certain periods of
time. The remote driver receives a
video stream taken from the cameras
on the vehicle. The real-time video
provides elaborate perception of the
environment to the driver so that the
diver can operate the vehicle as if
he/she is personally inside the vehicle.
Advanced video technology (e.g. VR)
could further improve the experience
of a remote driver.
Summary of requirements
Automotive use cases have huge business
potential, however, they also impose
stringent performance requirements for the
mobile system. Support of the CV2X
requirements has been introduced for LTE in
3GPP Release 14 [14], while with regards to
5G, the plan is to develop technical solutions
as part of Release 16 is expected to be
completed at the end of 2019. In parallel to
the 3GPP activity, 5GAA is working on
gathering relevant use cases and
requirements working directly with automotive
industry partners. The scale of the challenge
should not be underestimated: to support
advanced driving services, the mobile system
should provide ultra-high reliable and low
latency communication between the vehicles
and the network, as well as between vehicles
and vehicles, exceptional mobility
management performance and seamless
service continuity even when moving at high
speed. Furthermore, the mobile system
should provide “predictive QoS”, that is
inform the vehicle of changes in the quality of
the connectivity going to be provided in the
future so that the vehicle could decide to
switch from autonomous driving mode to
manual driving mode. Examples of factors
that the network may use to predict future
quality of service include weather conditions,
road situation, network availability at the
position where the vehicle is travelling etc.
5G may have geographic coverage
limitations (especially at the initial deployment
stage). It is therefore useful for 5GAA and
other relevant industry associations to specify
vehicle (or the device) behaviour in case the
network performance is not going to meet the
minimum requirements.
Energy
Overview
The energy vertical has some very specific
requirements on the supporting
communication solutions that go beyond what
current LTE can provide. The business
potential of introducing 5G Network Slicing in
the energy domain is exceptionally high, as it
is expected to provide the necessary support
not only to the critical machine type
communication (MTC) applications of energy
grid protection and control, but also to the
massive volume of MTC type applications of
the emerging smart metering.
Use Cases
Energy use cases have been investigated for
instance by 5GPPP[15] and 3GPP[16], as
well as by vertical customers themselves [32]
. Some typical use cases are briefly
explained in the following.
Smart grid
Due to the urgent demands of smart grids,
an efficient and reliable communication
26
network solution is expected. The
backbone network domain is a typical
network in which terminals are in the high
and extra high voltage area. The terminals
of a backhaul network are in the medium
voltage area. As for the access network,
the end points are in the low voltage area.
A large amount of growing demand
happens in the medium-voltage and low-
voltage domains, which are secondary
substations and distributed energy
resources, between primary substations
and secondary substations. At this
moment, there is a lack of energy
measurement and communications
system between substations. 5G Network
Slicing could be an economical and
efficient wireless solution compared with a
traditional fibre-based communication
system.
Micro-grids
Micro-grids consist of a set of micro-
power, load, energy storage and control
devices. It could operate in both grid-
connected mode and island mode. Micro-
grids will play a significant role in the
future electricity smart grid architecture
and the associated control network. All
Micro-grid elements need extensive
exchange of signalling between each
other. 5G Network Slicing could provide
an economical and efficient way to
support the communication needs.
Smart meters and aggregator gateways
Future power terminals are expected to
supply frequent measurements. This
evolution leads to the requirement on
future networks to carry short data
packages from thousands of users. The
data will enable near real-time
optimisation of sections of the low and
medium voltage infrastructures. This
optimisation will be particularly beneficial
for utilities to better serve customers
(residential and business) in densely
populated areas where 5G Network
Slicing is expected to become available
first.
Electricity traffic scheduling
Power supply and delivery play a more
and more significant role in modern life,
commercial and industrial operation.
Power outages as a result cause
significant economic damage both to the
power company and consumers. At
present fault location is performed by
means of fibre and Wi-Fi, which results in
Bandwidth 1 kbps per residential user.
E2E Latency (upper
bound)
<5ms between the primary substations and towards the
control centre.
Packet-loss
< 10-9 (which is more demanding for high/extra high
voltage than for medium voltage applications. This is
the end to end packet loss including the device.).
Availability >99,999% equal to about 5 minutes downtime per year.
Failure Convergence
Time
Seamless failover required, i.e. no loss of information in
case of a failure while keeping real-time delivery of the
information (i.e. within a small number of milliseconds).
Handling of crisis
situations
(Surviving long power downtimes on a large scale,
assuring black start capability): mandatory.
Connection density > 1000 km2
Table 3: summary of requirements for Energy industry
27
high latency and low level of reliability. 5G
Network Slicing could play an important
part in fault location and fault isolation by
transmitting critical monitored data,
controlling signals among feeders, bar
switches and automation control servers,.
5G URLLC Network Slicing has the low-
latency and high-reliability characteristics
needed to promptly and precisely respond
to power outages. Furthermore, by
building a precise map of power
consumption it is possible to improve the
traffic scheduling and production.
Summary of requirement
Table 3 provides, for reference a set of
requirements for common energy industry
services.
Healthcare
Overview
The health & wellness industry is about to
change dramatically thanks to the availability
and usage of electronic processes and
communication technology. This change will
be driven by demographical changes and the
resulting growing costs as well as the
growing demand for flexible and individual
treatment.
Use cases
The health & wellness use cases are very
diverse with many sub-use cases. The
following list aims to briefly summarise the
most important ones. More details to some of
these use cases can be found in [17]:
Hospitals
Deals with processes in and around a
hospital. The most prominent examples
are medical device tracking, emergency
communication, (real-time) data
availability and exchange and data
transfer.
Rehabs and care homes
Processes in and around rehabs and care
homes. Here, the most important use
case is assisted living.
Health and wellness monitoring
Tracking health-relevant indicators using
various types of sensors. This data can be
collected at a central office where it can
be analysed by automated data analytics
processes as well as human experts.
Remote healthcare
Live bidirectional video for more efficient
consultation, diagnosis, treatment,
monitoring as well as for assisted surgery.
Remote surgery
In this use case medical devices are
controlled remotely allowing off site
surgeons to operate on a patient. This is
probably the use case with the most
demanding requirements in terms of
latency, reliability and guaranteed quality
of service. It is unlikely that the early
deployments of 5G networks will have
sufficient capabilities to support the
functionality required by remote surgery.
Summary of requirements
The requirements for the health & wellness
industry need to be further analysed. With the
exception of remote surgery, it seems
reasonable to assume that the requirements
on the network are not very stringent,
therefore it should be possible to support
most of the use cases when the 5G network
is deployed.
Availability is most important requirement for
these use cases, followed by latency and
then throughput.
From a functional point of view, many of the
mentioned use cases are likely to require
precise positioning solutions that also work
when the device is indoor, strong security
and both cloud and edge computing
capabilities.
28
Industry 4.0, Manufacturing
Overview
The “Fourth Industrial Revolution” or Industry
4.0 is set to fundamentally change the
manufacturing industry. The main drivers are
improvements in terms of flexibility,
versatility, resource efficiency, cost efficiency,
worker support, and quality of industrial
production and logistics.
To achieve the required flexibility, wireless
connectivity (as a substitute for the wire-
bound connectivity available in today’s
factories) is essential. Generally speaking,
Industry 4.0 use cases have very demanding
requirements, e.g. in terms of latency,
synchronicity and availability.
Use cases
In [18] the Industry 4.0 use cases have been
grouped into five different application areas.
These are: Factory automation, Process
automation, Human-machine interfaces
(HMIs) & production IT, Logistics and
warehousing and Monitoring and
maintenance.
These application areas are composed of a
set of basic use cases. These use cases are
briefly summarised below:
Augmented reality
Augmented reality optimally supports
shop floor workers, for instance, in tasks
like: monitoring of processes and
production flows, step-by-step instructions
for specific tasks (e.g. in manual
assembly workplaces), ad-hoc support
from a remote expert (e.g. maintenance or
service tasks). It is expected that the AR
devices have minimum capabilities and
that complex functions are executed at
the edge cloud.
Control-to-control (C2C)
Control-to-control (C2C) communication
refers to the communication between
different industrial controllers. For higher
flexibility it is expected that this
communication is wireless. Depending on
the concrete scenarios, these C2C
systems typically have very challenging
requirements on the communication
service.
Motion control
Motion control controls moving and/or
rotating parts of machines. Wireless
communication is well suited for the
control of components, which move
and/or rotate.
Mobile robots and mobile platforms
Mobile robots and mobile platforms
perform activities like assistance in work
steps and transport of goods, materials
and other objects and can have a large
mobility within the industrial environment.
Mobile Control Panels with Safety
Functions
Mobile control panels with safety functions
are devices for the interaction between
people and production machinery as well
as for the interaction with moving devices,
e.g. for configuring, monitoring,
debugging, controlling and maintaining
machines, robots, cranes or complete
production lines. Optionally these panels
are equipped with an emergency stop
button.
Closed-loop control
Closed-loop control in which several
sensors are installed in a plant and each
sensor performs continuous
measurements. The measurement data is
transported to a controller, which takes a
decision to set actuators.
Process monitoring
Process monitoring in which several
sensors are installed in the plant to give
29
insight into process or environmental
conditions or inventory of material. The
data is transported to displays for
observation and/or to databases for
registration and trending.
Plant asset management
Plant asset management that is required
to keep a plant running. It is essential that
the assets, such as pumps, valves,
heaters, instruments, etc., are maintained.
Timely recognition of any degradation and
continuous self-diagnosis of components
are used to support and plan
maintenance. Remote software updates
enhance and adapt the components to
changing conditions and advances in
technology.
Summary of requirements
Regarding performance, the most demanding
requirements of the Industry 4.0 use cases
are latency, reliability, device synchronicity,
data rates, seamless mobility and energy
efficiency. The most important functional
requirements are plug-and-play support,
positioning and local edge cloud support. For
Industry applications, safety and security
requirements are very critical and different
from the other industries. Malfunction of the
communication and potential external attacks
via hacking can cause significant damage of
production process even employee health.
Moreover, 5G industrial solutions also need
to consider how to seamlessly integrate the
legacy communication methods used in the
industrial environment.
Campus networks are well suited to satisfy
the local coverage requirements of the use
cases. Due to the very stringent requirements
of these use cases, especially in terms of
availability and reliability, some of the factory
owners believe that arrangements such as
exclusive access to spectrum and a private
network operated by the owner of the factory
will be needed.
The high throughput needed for some use
cases, for example to support AR, will require
a large bandwidth that may only be available
in the above 6GHz spectrum bands.
Internet of Things for Low Power Wide Area
Applications
Overview
The Internet of Things (IoT) describes the
coordination of multiple machines, devices
and appliances connected to the Internet
through multiple networks [19]. The IoT
encompasses a huge variety of use cases
including some of the ones already described
above. This section will cover use cases
relevant for Low Power Wide Area (LPWA)
applications, where the goal is to improve
power efficiency, coverage, total cost of
ownership and sustainability. Areas like
agriculture and environment, consumer,
industrial, logistics, smart building, smart city
and utilities [20] are addressed.
There are several use cases, some of them
have been also included in the other vertical
industries (i.e. road warnings, smart meters,
etc.) since they apply to a huge variety of
sectors. In the following, only the most
important use cases not address by other
vertical industries are briefly described.
Use cases
The following list summarises the most
important use cases:
Asset Tracking and monitoring
Being able to track and monitor assets is
a very attractive use case because it
allows to optimise the operational
aspects, which differs depending on the
type of asset. There are available
30
examples of assets tracking that range
from suitcases to pallets or containers.
Waste management
To minimise both costs and the impact on
the environment, city administrations are
looking to make waste collection smarter.
Sensing when residential bins require to
be collected improves operational cost. In
addition, there could be a variety of
sensors installed in bins, like for detecting
the emission of gasses or fire.
Smart parking
Sensors are used to collect data about the
occupancy of each parking space are
becoming a very prominent use case for
IoT. This solution not only improves traffic
congestion, but it can also improve
revenue collection where parking is paid
for and improve usage of parking spaces
where drivers can be directed to a free
space.
Smart manhole
Smart manhole can be used to detect
leakage in the infrastructure, both for
water and gas. Smart pipeline monitoring
can be deployed by utilising a variety of
sensors like pressure, water flow, gas
detector and acoustic, which report
constant measurements to the control
room.
Water metering
Involves the use of a flow sensor that
allows measuring the water consumption.
Water meters tend to not have any access
to mains power and with an average
lifetime that exceeds 10 years and in
some cases reaches up to 16 years.
Gas Metering
Gas metering is very similar to the water
metering use case. Gas meters have
stricter requirements on the power
consumption in order to prevent
explosions. Also for this use case, the
expected lifetime for the meter is normally
set by national regulations, for example in
Europe it is generally for 15 years or
above, while in China it is around 10
years.
Summary of requirements
The requirements for LPWA have been
analysed in quite some detail already. From
the above use cases, the following
requirements can be highlighted:
A small amount of information is generally
exchanged per transmission, low
throughput required.
Latency tolerant, better latency
performances are expected for use cases
like asset tracking and parking, but still
acceptable in the range of a couple of
seconds.
Coverage is particularly important, for
some use cases coverage is more
important indoor and for other it is
outdoor, but it is a constant requirement
for all the use cases.
Power consumption, all use cases
described above do not have access to
mains power and therefore require
mechanisms to reduce as much as
possible the battery consumption of the
devices.
Mobility is only relevant for asset tracking
and marginally for waste management.
IoT is a good example in which most of the
use cases can be served with available
technology, such as 2G, NB-IoT, etc.
However, these are networks built and
operated independently from each other.
Serving the IoT use cases over 5G and
Network Slicing technology has the potential
to greatly reduce operational costs, increase
efficiency and leverage additional capabilities
an operator can offer.
31
Public Safety
Overview
Several countries are considering updating
their existing PPDR networks to take
advantage of the advancements in mobile
technology and capabilities such as high
speed data transfer that are underdeveloped
in the current mainstream systems in use
such as TETRA and P25. Adopting the new
technology will also unlock economies of
scale and the introduction of innovative
solutions as they become available for
commercial services.
Countries such as the United States, South
Korea and the UK are spearheading the
introduction of LTE-based PPDR networks
and the trend seems to point to a deployment
of such networks as an overlay of existing
commercial networks. A public safety network
is therefore conceptually consistent with a
network slice and due to the performance,
availability, isolation properties it is natural to
expect that it will be in a future deployed in a
network slice.
Use Cases
The following use cases have been identified
for public safety enabled mobile networks:
Mission Critical Push-To-Talk
As for the legacy TETRA and P25
networks, users of public safety networks
need to be able to communicate in groups
characterised often by strict hierarchy and
facilitated by powerful floor control
mechanism. The enablement of MCPTT is
underpinned by the following
requirements:
Very low end-to-end talk setup time;
below 100ms
Very high number of users in one cell;
ca. 6000
High quality voice, incl. noise
cancellation
Group call capabilities
Broadcast call capabilities
Pre-emption and prioritisation for
specific services, like Emergency Call
Symmetric utilisation
End-to-End encryption; 256 bit or
higher
Interface to narrowband standards,
(e.g. TETRA)
Mission Critical Data
Data communications to allow the
exchange of text messages, files and
images between public safety officers
can greately enhance the efficiency of
their operations when compared to
using only voice media. The following
requirements are expected to be met:
Real-time capability with very low
delay and jitter, due to automatic
analysis and overlaying during
simultaneous usage of different
sources
High bandwidth in one network cell; up
to 5 Gbit/s
Pre-emption and prioritisation for
specific services
Symmetric utilisation
End-to-End encryption; 256 bit or
higher
Interface to narrowband standards,
like TETRA
Mission Critical Video
Real-time capability with very low delay
and jitter, due to automatic analysis and
overlaying during simultaneous usage of
different sources
High bandwidth in one network cell; up
to 5 Gbit/s for video signals with a
minimum resolution of 1080p60
Pre-emption and prioritisation for
specific services
32
Asymmetric utilisation; higher uplink
capacity than downlink
End-to-End encryption 256 bit or
higher
Interface to narrowband standards,
like TETRA
Massive Mission Critical IoT
Access to the ever-growing number of IoT
devices such as security cameras,
drones, smoke detectors, health monitors
will provide unvalued support to the
operations of public safety agencies, it is
therefore required that a network can
support communication with these
devices in a secure and reliable manner.
The following requirements apply to IoT
devices used for mission critical
operations:
Real-time capability with very low
delay and jitter, due to automatic
analysis and overlaying during
simultaneous usage of different
sources
Pre-emption and prioritisation for
specific services;
Asymmetric utilisation
End-to-End encryption; 256 bit or
higher
Interface to narrowband standards,
like TETRA
Summary of Requirements
Many commonalities emerge from the
analysis of the requirements of the public
safety use cases. Some of the requirements,
such as backwards compatibility towards
legacy PPDR communication systems, can
probably be realised at application level via
some interworking function. Security, which
stands out as a main requirement for public
safety use cases, could be provided at
application level (e.g. end to end encryption),
however it seems a better choice to utilise the
capability of implementing different security
models as well as isolation properties that
Network Slicing allows to fulfil the security
requirement of the public safety use cases.
The network slice will also need to be
designed to provide capabilities for pre-
emption and prioritisation of mission critical
traffic, radio bearers with high bandwidth for
mission critical data or with low delay and
jitter for MCPTT and Mission Critical Video.
Smart Cities
Overview
The concept of smart cities envisages the
deployment, management, usage and
maintenance of city’s assets using ICT. The
goal is to improve efficiency, sustainability
and to address the requirements coming from
changing demographics. Thereby, areas like
public transport, infrastructure, utilities, public
safety, etc. are addressed.
The smart cities use cases are very diverse
and many of them have already been
addressed within other vertical industries
(e.g. waste management, smart parking,
smart bike sharing, energy grid, etc.). The
following section focuses on the most
important use cases that are not already
addressed in this paper as part of other by
other vertical industries are briefly explained.
Use cases
Intelligent lighting
Intelligent lighting refers to lighting
networks in which, for instance, lights can
be turned on in a formation on-demand. In
addition, maintenance gains can be
improved through regular status reports.
However, the gains are expected to be
relatively low. Therefore, the bundling with
other services is needed, e.g. for
monitoring, reporting, and coordination
tasks as well potential sites for small cells
or roadside units (for V2X services).
33
Public safety
Public safety networks open new
opportunities to detect and fight crime.
Examples could be secure and reliable
communication networks for security
service and law enforcement as well as
the real-time analysis of data coming from
different sources like cameras, sensors,
etc.
Emergency service management
Emergency service management is about
efficient and target-oriented
communication in disaster (natural as well
as man-made) scenarios. Herby,
warnings to the people as well as
communication to non-human entities
(e.g. shutting down elevators, etc.) is
envisioned.
Summary of requirements
The requirements for the smart cities industry
have not yet been analysed in detail. For
most of the use cases and requirements no
concrete values but indications are available.
However, it turns out, that most of the use
cases are not very demanding e.g. compared
to other industries.
Regarding the performance requirements, a
very high density of devices as well as the
high availability/coverage and high data rate
are the most important requirements.
From the functional requirements energy
efficient operations, especially for sensors
and security are most important.
34
5 Industry
Recommendations and
Reshaping the Network
Slicing Concept
35
Industry recommendations and reshaping the Network
Slicing concept
This chapter aims to conclude the findings
from the service requirements analysis
discussed in section 4. Rather than providing
details about the collected service
requirements, this chapter provides
conclusions and recommendations derived
thereof. Recommendations towards cross-
standardisation collaboration are expressed
and the concept of the generic networking
slicing template is briefly introduced. In
addition the concept of Network Slicing is
reshaped for the purpose of clarifying some
common misunderstandings.
Requirements analysis
Use case clustering
This section provides a descriptive
comparison outlining similarities and
differences in use case requirements. In
the future this clustering can be extended
allowing identification of required service
slice types and their characteristics.
In the literature future use cases are
normally clustered based on their
performance requirements into:
Enhanced mobile broadband (eMBB)
Ultra-reliable low latency
communications (URLLC)
Massive machine type
communications (mMTC)
eMBB in contrast to MBB provides
improved data rates, capacity and
coverage. URLLC refers to critical types
of communication supporting very low
latency, high reliability as well as small to
medium data rates. mMTC supports the
IoT use cases with scenarios in which a
very large number of (millions to billions)
of small devices have to be connected
efficiently, e.g. in an energy efficient way.
The use cases discussed in this
document can be also grouped into these
three main groups. An overview for non-
IoT use cases is provided in Figure 7 in
which use cases are roughly categorised
based on their throughput, latency and
Figure 7: Use case clustering
36
availability requirements1. However,
based on the performance requirements
at least two additional service types can
be identified:
eMBBLLC use cases have high
requirements on latency and
bandwidth at the same time.
Availability requirements are in most of
the cases a bit more relaxed
compared with URLLC use cases.
Examples are augmented reality and
virtual reality use cases.
mMTC/URLLC use cases comprise
sensors sending data which needs to
be delivered with a very low latency
and a high reliability. One example is
the motion control use case of Industry
4.0.
Based on the previous findings one of the
most important steps is to identify service
slice types required to serve the different
use cases. These service types could be
selected based on performance,
functional or operational requirements.
How to do it exactly is subject to future
discussions.
Performance requirements
This section highlights the most important
performance requirements requested by
the vertical industries. Requirements are
explained and initial ideas are discussed
on how to address these requirements.
Very tight synchronization
Some use cases have stringent
requirements in terms of synchronicity
of communication devices. Examples
are cooperative driving use cases from
1 It should be noted that the clustering should be seen as
indicative. Where available concrete values have been
considered in case ranges of values were available the worth
the automotive industry or motion-
control and control-to-control use
cases from the Industry 4.0 vertical.
For the latter ones, values of smaller
than 1 microsecond are required.
Today these requirements are met by
the deploying cable
connections/networks as like industrial
Ethernet systems or fieldbuses. These
networks are normally closed solutions
by a single vendor in which all the
equipment is perfectly aligned. Based
on standards like IEEE 802.1 AS [29]
or IEEE 1588 [30] very high
synchronicity can be achieved in the
networks.
When thinking about 5G networks it is
the expectation of the vertical
industries that these very demanding
synchronicity requirements can also
be provided by wireless and cellular
network. However, especially in
wireless scenarios and for moving
communication partners this is not an
easy task and it requires more detailed
investigations.
Cyclic traffic
Some applications rely on
cyclic/deterministic traffic. Cyclic traffic
is traffic with very regular traffic
patterns, e.g. inter-packet delay (X) –
see Figure 8.
Relevant use cases are, for instance,
Voice over IP (e.g. 20ms inter packet
delay), motion control (e.g. < 0.5ms
case was assumed and where no values were available and
educated guess was done.
Figure 8: Cyclic traffic
time
X X X…..…..
37
inter packet delay), control-to-control
(e.g. 4ms inter packet delay), etc.
The network slice must be able to
deliver this type of cyclic traffic with
minimum impact on the inter-packet
delay. Especially in the RAN, this
requires a special treatment. The pre-
scheduler feature available in LTE
since Release 8 might be a good
candidate as it allows sending traffic in
specific intervals by assigning
resources in regular intervals. This
reduces the delay and jitter as well as
signaling overhead which improves
the efficiency. However, more analysis
is required.
Functional requirements
This section highlights the important
functional requirements derived from the
service requirements discussions. Some
basic functions that operators could
provide for industry customers include
authentication, firewall, identity
management, etc. Upon these
fundamental requirements, a number of
new functional requirements are also
indicated by variant vertical industry use
cases.
Security
Security is a general concern for all
industries, because of the damage
that security leaks may cause.
However, the security requirements
may be very different from industry to
industry. For instance, the method to
provide secure connectivity for millions
of low-cost sensors could be very
different to the mechanisms used to
provide security for public safety
services. Appropriate technical
solutions should be defined to meet
these requirements and comply with
the requirements and solutions
defined by 3GPP [28].
Isolation
Isolation refers to the degree of
resource sharing that could be
tolerated by the industry partner.
Some customers may not mind to
share network resource with other
customers, but would require isolation
for the computing resource. Sensitivity
Figure 9: Positioning accuracy and energy efficiency indoor and outdoor
38
or criticality of the processes used by
some customers may on the other
hand lead to the requirement to only
want to share the physical site like
base stations, but use dedicated
spectrum. Network Slicing should be
able to be configured with different
levels of isolation to satisfy the
customers’ needs.
More details about isolation are
discussed in the 0 on page 44
Positioning
Many use cases have a strong
demand for the capability of
positioning (geo-localisation) devices.
Different customers may have different
requirements in terms of accuracy,
energy efficiency, indoor/outdoor
support, and cost, etc. For some of the
use cases, positioning techniques will
have to work reliably under
challenging conditions, e.g. deep
indoors.
Figure 9 graphically represents the
positioning requirements in terms of
accuracy and energy efficiency for a
number of use cases.
For automotive use cases (A), e.g.
automated cooperative driving like
short distance platooning, high
precision positioning feature with an
accuracy of a few centimetres is
expected to be supported by the
mobile network to ensure the vehicle-
to-vehicle/network communication can
be used even when GPS is not
available, such as when moving
through very dense urban scenarios or
in bad weather condition.
For manufacturing use cases (B), high
precision indoor positioning with an
accuracy of <1m is required to support
for instance mobile control panel with
safety functions.
For asset tracking type of use cases
(C), like.g. container tracking in
Transport & logistics sector, it is often
sufficient to provide a positioning
solution that works outdoors and with
relatively low accuracy. On the other
hand, as mobile devices may be
battery operated, the energy efficiency
feature of the positioning solution
plays an important role. In addition, in
many of these use cases the
positioning is expected to work
globally across different networks and
countries.
It should be noted that for some use
cases, such as many IoT use cases,
GPS or other Global Navigation
Satellite Systems (GNSS) are not an
option either because of the high
energy efficiency consumption or
because simple devices are not
equipped with the suitable receiver.
For other use cases, e.g. automotive,
GNSS is a suitable positioning solution
for most of the times although there
are a number of situations where this
is not accessible e.g. tunnels, indoors,
etc. Hence, it is beneficial to combine
the advantages of these systems with
the positioning capabilities of 5G to
provide a solution that meets the
customer scenarios.
Delay tolerance
To support certain vertical industry use
cases, mobile system should be able
to provide guaranteed SLA that is
agreed with the customers. For
instance, certain traffic flows should
reach the end user within certain
latency boundary. At the same time,
there are use cases that are less
sensitive to delay variations giving the
mobile system some level of flexibility
in scheduling traffic. For instance, in
automotive industry, (non-critical)
software/firmware update could be
39
deprioritised and delivered when traffic
is low such as during off-peak hours.
Predictive QoS
Predictive QoS is an important feature
allowing operators to inform the
service in advance about a quality
drop. Predictive QoS can be applied to
various KPIs, e.g. coverage,
throughput, latency, etc.
For instance, when a vehicle is moving
along a path under autonomous
driving mode, the mobile system may
foresee performance drop (e.g. due to
channel condition or system
congestion situation). In this case, the
mobile system could as a precaution
take corrective actions e.g. resource
rescheduling (scaling up/out), or
simply inform the vehicle about the
potential performance drop. Such
information could be evaluated by the
vehicle itself in order to make decision
to switch from autonomous driving
mode to manual driving mode to avoid
potential danger. The higher accuracy
a system’s predictive QoS provisioning
capability is, the higher possibility for
the system to support industry use
cases with critical system performance
requirements.
To find out how predictive QoS works
under realistic conditions and which
accuracies can be achieved more
analysis and field trials are required.
Operational requirements
In previous generation mobile systems,
operator’s infrastructure as well as the
provided network services have always be
presented as a black box for the OTT or
enterprise customers. Only few APIs are
provided by the network to the service
providers including making and receiving
voice calls, sending SMS/MMS, the Service
Capability Exposure Function (SCEF) that
provides a means to securely expose the
services and capabilities provided by 3GPP
network interfaces, etc.
However, these APIs might not be enough for
some of the future use cases or future
customers. To better support vertical
industries to deliver their services to their own
end users via slicing method, under specific
regulatory regimes it may be possible to open
up certain operational capabilities of the
mobile system towards the vertical industries.
Operators could provide a set of APIs to
expose such capabilities towards vertical
customers. This does not only provide
convenience for verticals to manage their
own service, but also enrich the Network
Slicing business model for telco industry.
According to the business incentive as well
as service operation methods, vertical
customers have very different perspectives in
terms of required operational capabilities.
Monitoring capability
Monitoring capability is one of the
fundamental operational requirements, for
instance monitoring traffic characteristic
and performance (e.g., data rate, packet
drop, and latency), end user’s
geographical distribution, etc. Moreover,
monitoring itself also contains many
subcategory according to the add-on
requirements, for instance, monitoring
granularity (e.g. per second, per hour,
etc.), per session/user/slice instance
based monitoring, etc.
Limited control capability
Vertical customers could use APIs
provided by the operators to control
network service. For instance, instead of
simply using the network service, vertical
customers could decide where to locate
the network functions. Moreover, verticals
could also decide how to integrate their
own applications with the other network
functions to run together within the
network slice instance provided for them.
40
Configuration capability
Configuration refers to the capability to
allow vertical customers to adjust and
modify the network functions as well as
underlying resources within the network
slice instance provided for them.
Full operation capability
Full operation is the highest operational
capability that could be provided via
slicing method. Within the provided
network slice instance, vertical customers
could have full control of the network
service operation. Therefore, they are
self-responsible for the network service
assurance and maintenance within the
limit set by the network operators. In
particular network access should remain
under the control of the operator. The
impact and the application of the
regulatory obligations will have to be
studied in this approach.
This might be very helpful for the
customer in case of networks problems or
outages for instance by configuring
backup paths.
Coverage requirements
Vertical requirements very much
differentiate in the type of coverage they
require. Three main groups have been
identified: Local, nationwide and global
coverage. More details are provided in the
following.
Local coverage
Local coverage is requested mostly by
Industry 4.0 and healthcare industries
in which a very specific geographical
area needs to be covered, e.g. an
industry area or a hospital.
Local coverage requirements could be
met by a network slice (as a kind of
VPN based on an existing public
network) restricted to a small number
of base stations or by so called
campus networks. A campus network
is a dedicated physical network
installed in the premises of the
enterprise. This makes campus
networks very flexible (i.e. build out
independent of an already existing
public network) as they are limited to a
small number of devices, which allows
the rollout of new hardware fast and
efficiently.
Typically for local coverage a single
administrative domain is sufficient to
serve the customer or use case.
Nationwide coverage
Nationwide coverage requested by
industries, which do not need
coverage outside a specific country.
Examples are utilities industries,
government services and smart cities.
For this type of coverage, typically a
single administrative domain (e.g. a
single network operator) is sufficient.
For very demanding use cases,
national roaming could be envisioned
in order to meet for instance the
availability and coverage
requirements. According to NGMN
[26] this is the “business vertical”
scenario.
Global coverage
Global coverage is required by
companies and industries with
worldwide points of presence or which
are selling their products globally. The
most obvious industry requiring global
coverage is automotive. Other
examples are transport & logistics use
cases. Providing global coverage
normally means that network services
need to be provided across multiple
administrative domains. According to
NGMN [26] this is the “roaming”
scenario.
For this type of coverage, the network
slices used to serve customers need
to be made available also when the
device is outside its home network.
41
Several solutions are discussed in
literature on how to achieve this.
Three of the most prominent
approaches are briefly introduced: (1)
the visited network could provide to
the roaming user a network slice with
equivalent functionality of the slice
used in the home network, e.g. using
standardised slices. (2) the home
network may export the blueprint of a
custom network slice used by a user
so that it can be instantiated and
administered by the visited mobile
network operator. (3) the home
network may extend the slice into the
visited network, provided it has
authorisation from the visited network
to control the resources.
Generic Network Slicing template
As discussed in this section, a network slice
has different characteristics or attributes
which can be roughly grouped into three
categories: performance, functional and
operational.
These attributes need to be specified and
quantified in some way, so that:
the slice buyer knows what to expect and
to be sure that the service requirements
can be met
the slice provider knows what to deliver
there is a sound basis for a
SLA/contractual agreement between the
Slice provider and buyer
UEs can roam onto another networks, but
maintain access to the services that
require particular network capabilities e.g.
by allowing a home operator to buy a
Slice in a visited domain
The general set of attributes that characterise
a slice (in terms of performance, functionality
and operation) should be the same for any
slice implementation, regardless of the values
associated to those attributes.
NOTE: some attributes may be void or not
applicable to some slice implementations
It is highly beneficial to achieve a common
method the industry can refer to in order to
describe the characteristics of any slide – that
is what we refer to as a “Generic Slice
Template (GST)”. Every Slice can be fully
described by allocating values (or ranges of
values) to each relevant attribute in the GST.
One of the main areas the industry should
focus is to agree a definition of a Generic
Slice Template, with all the KPI names, units
and granularity, but no actual values. This
would allow to unlock all the benefits
described above. For example, an operator
may easily communicate the expected
behaviour of a network slice provided to a
customer when such customer moves to a
visited network allowing the visited network to
decide what mappings would be needed
between slices in home and visited networks.
This generic template would not be specific to
a vertical, and so would be useable by any
future 5G services. When deploying or
offering a slice to a customer, the “contract”
would specify the values that each parameter
would take for a given slice instance. Some
examples of parameters, or ranges for
parameters may be given to communicate
the intent.
The generic slice template serves as a
baseline for the definition of service/slice
types by specifying the values (or range of
values) for each of those attributes. The GST
also serves as baseline for slices only
negotiated between the operators and its
customer.
42
Cross-Standardisation Collaboration
Recommendations
E2E Network Slicing contains different
technical domains as discussed in Section 3
(e.g. AN, CN, TN, Cloud), and many SDOs
are working in parallel to provide slicing
solution under their area of competence. The
technical content is as a result fragmented. In
order to form an E2E solution, significant
work is required in terms of cross-SDO
cooperation and coordination, which will be
explained in the following sections.
Coordination from System Architecture
Perspective
The overall 5G system architecture is under
the development of 3GPP SA2 Working
Group, for whom Network Slicing is one of
the fundamental design feature of the 5G
system. To enable E2E connectivity of
Network Slicing across multiple technical
domains, potential coordination between
3GPP and other SDOs is desired. For
instance, in an FMC (Fixed Mobile
Convergence) scenario, part of the E2E slice
may be provided using non-3GPP access
such as WLAN or fixed access defined in
BBF), therefore combined efforts between
3GPP SA2 and BBF may be required to
guarantee the correct deployment.
When considering slices that should fulfil use
cases with low latency requirements that can
be provided using Multi-access Edge
Computing (MEC) coordination between
3GPP SA2 and ETSI MEC [21] would be
required. For instance, some information
such as application instances locations, traffic
rules, identification of traffic that needs to be
offloaded to the edge needs to be
exchanged between 3GPP network and ETSI
MEC system for the purpose of selection
and/or reselection of the User Plane Function
(UPF) and potentially to relocate the
application. There is an ongoing work item in
ETSI ISG MEC named “MEC support for
Network Slicing” [24], which focuses on
identifying the necessary support provided by
Multi-access Edge Computing for Network
Slicing and, in addition, how the orchestration
of resources and services from multiple
administrative domains could facilitate that.
This could be a potential candidate for cross-
SDO coordination for MEC topic.
Coordination from Management
Perspective
Transport Network Aspect
As shown in Figure 6 of Section 3,
management aspect is considered
separately in different SDOs based on the
targeted technical domain. 3GPP SA5
Working Group is responsible for the life
cycle management of slices (e.g. creation,
run-time assurance, etc.) in the 3GPP
realm that contains the RAN and CN parts
of the slices. Transport network, which is
one essential part in E2E Network Slicing
solution, is not in 3GPP scope. TN slicing
management architecture is instead
defined by BBF. To enable E2E
management of a slice, it is essential to
align BBF and 3GPP SA5, in terms of
terminology, architecture, interfaces, etc.
Moreover, interoperation between 3GPP
and BBF-defined slicing management
entities needs to be in place. For instance,
3GPP SA5-defined slicing management
system needs to send transmission link
requirements to BBF-defined TN slicing
management system. Vice versa, 3GPP
slicing management system needs to
retrieve TN information to establish the
transmission link between 3GPP network
function entities. The northbound interface
that is used for such information
exchange should be aligned between BBF
43
and 3GPP, and IETF could be used to
standardise the protocol of the
northbound interface according to the
requirements collected by BBF.
Virtualisation Aspect
Virtualisation and cloudification have been
deployed in 4G cores already. This
technologies will continue to be the major
features in the 5G new core (5GC)
design. 5GC network functions are
modularised (so that they can be
executed as software in a datacentre) and
interconnected with each other using
service based interfaces. Hence, to
deploy and manage slices in a virtualised
network environment, 3GPP SA5 should
coordinate with ETSI ISG NFV.
Coordination from Security Perspective
Security is a built-in feature of the overall
system architecture that supports Network
Slicing in order to provide the fundamental
guarantee of the telecom services as well as
the specific requirements of the safety-critical
applications and services. Way beyond the
slice-internal and slice-specific security
provisions and the access control to the
slicing subsystem, slice security should be
also systematically designed to empower
secure and reliable E2E slice life cycle
management.
E2E slicing security can be implemented at
different levels and dimensions, e.g. from
communication method (e.g. authentication)
to isolation method (e.g. soft isolation based
on virtualisation or hard isolation based on
physical entity). Security related work is
carried out in multiple SDOs in parallel as
well. 3GPP SA3 works on the
security/isolation method for Network Slicing,
including the definition of overall
implementation, strategies, standard
interfaces and protocols etc., but as in the
case of slice management, it does not cover
transport as well as virtualisation aspects.
Hence, the coordination between 3GPP SA3
and ETSI ISG NFV, BBF, and IETF are
necessary for an overall E2E security.
Coordination between SDOs and Open
Source
Other than the work from SDOs, open source
is fast becoming more and more prominent in
the scope of 5G. For example, Open Network
Automation Platform (ONAP), an initiative
from operators in cooperation with the Linux
Foundation, is an open source software
platform that delivers capabilities for the
design, creation, orchestration, monitoring,
and life cycle management of VNFs, carrier-
scale Software Defined Networks (SDNs) as
well as higher-level services [22]. Open
Source MANO (OSM) is another operator
initiative from ETSI community, which aims to
deliver production-quality open source
Management and Orchestration (MANO)
stack aligned with ETSI NFV [23]. Currently,
there is a lacks of clear understanding of the
relationship between standardisation and
open source especially at the management
domain as well as competition between open
source initiatives.
As mentioned previously, ETSI’s Zero touch
network and Service Management (ZSM)[24]
aims to help in this area of coordination
among SDOs and open source solutions.
Part of its mandate is to analyse gaps and
help align approaches from these different
organisations. Currently, one of the work
items in ZSM is to develop an end to end
view of Network Slicing management looking
across domains, for example, 3GPP network
slice management and the ETSI NFV
virtualisation platforms.
44
Other recommendations
Additional recommendations are:
It turned out that for many KPIs, e.g. for
latency, different interpretations and
understanding throughout the industries
exist. Therefore, it is recommended to
specify/define relevant KPIs in order to
get a common understanding.
Requirements some industries pose on
5G and Network Slicing have been
derived from today’s use cases and
processes. However, today these
requirements are met using fixed line
network technologies as like industrial
Ethernet or optical fibres. These
requirements very often are too
demanding for cellular networks as 5G.
Therefore, discussions between
telecommunication and vertical industries
are required to cooperatively find
solutions for that.
Some use cases and industries have very
demanding availability and reliability
requirements. However, it is not well
understood how these requirements can
be met, e.g. which mechanisms can be
applied to improve availability and
reliability beyond state of the art networks.
It is recommended to have a deeper look
into these aspects.
It would be beneficial to standardise a
small set of globally available network
slices serving use cases that are of
interest in many countries and common to
different verticals and use cases.
Network Slicing Concept Evolution
Compared to 2015, when NGMN first
introduced Network Slicing concept, the
understanding of Network Slicing is much
more mature now. The fundamental system
architecture defined by 3GPP is capable to
support a terminal to use multiple slices
simultaneously. There are a number aspects
that need particular attention.
Operational isolation vs. network level
isolation
Figure 10: Different Levels of Network Isolation
45
“Isolation” is considered as one of the key
features that Network Slicing could provide.
Isolation could be understood from two
dimensions: (1) operational isolation, means
that vertical customers could have
independent monitoring, control,
configuration, or even full operation capability
of the network slice; (2) network level
isolation, means that vertical customers do
not share network function or resources with
the other customers. Network level isolation
also has different sub-categories, for
instance, shared RAN but isolated core, or
isolated RAN as well as core, etc.
Operators can provide operational isolation
without or with very weak network isolation.
For instance the system could use IDs to
differentiate the users belonging to different
tenants who share the infrastructure. One
example is NB-IoT, which can be treated as a
pre-configured network slice, with many
different IoT tenants sharing the same NB-
IoT network.
On overview of the different levels of network
isolation is provided in Figure 10. It is obvious
that the different levels of isolation will have
different cost. Most expensive mode will be
dedicated RAN (L0 or L1), which may only be
relevant for very few use cases.
It is essential to understand, how to let
vertical customers comprehend this feature
from business aspects and which kind of
isolation do they prefer to have.
From an operator’s perspective, isolation
requirements could be identified in different
ways. In some cases, it is the vertical
customers’ direct requirement, i.e. they
directly ask for operational isolation or
network level isolation. In other cases,
customers do not express any isolation
requirement, but the operator takes decisions
on isolation based on other requirements. For
instance, if a customer requests a service
level which is quite demanding, the operator
may assign dedicated resources in order to
ensure it can meet the service level
commitment, which in turn is implemented via
Network slice isolation solution.
Based on a survey with vertical customers,
the majority of those who have isolation
requirements, have direct requirements on
operational isolation instead of network level
isolation. There are also some specific
vertical industries that have clear network
level isolation requirements, e.g. public
safety, smart grid, for security and safety
purpose. Hence, their network slice(s) may
not share any network function and resource
with the other network slices. Such
customers are minority cases.
Not a network slice for each vertical
customer
Based also on the clarification of slice
isolation matter, it is not necessary to provide
an independently deployed E2E network slice
for each vertical customer, because this is
not their fundamental requirement. One
network slice could serve many vertical
customers simultaneously.
Number of slices supported by the 5G
system
This is one of the fundamental question about
Network Slicing. Potentially, the amount of
network slices could be very high under the
assumption of full system automation and
programmability. However, the amount of
slices should be designed based on the
capability of the underlying deployed
infrastructure, as well as the performance,
functional and operational requirements of
business customers (especially at the initial
phase of slice deployment). Moreover, it
ought to be an individual operator choice. It is
technically feasible to have roaming solutions
where the HPLMN and VPLMN have taken
different strategies in the number of deployed
network slices to serve the same customers.
Network slice deployment
Network Slicing deployment will benefit from
network automation. The automation in rolling
out network slices will be subject to normal
46
learning curve that new technologies are
commonly experiencing. So it can be
expected that the degree of automation of
deployment of a network slice will evolve over
time.
47
6
Regulatory Aspects
48
Regulatory aspects
Network Slicing will be a feature of 5G - this
is based upon software defined network and
network function virtualisation. In reviewing
Network Slicing, regulation should be
considered and views on key areas are
provided below. However, from a regulator’s
perspective they will want to understand
what, if anything changes from developments
in technology, and how this impacts
customers. These changes might impact and
/ or be constrained by the regulatory aspects
listed below. Further analysis is required to
understand the evolution of the regulatory
aspects to align with these changes.
Net Neutrality
While there is no single definition of net
neutrality, it is often used to refer to issues
concerning the optimisation of traffic over
networks.
Mobile network operators face unique
operational and technical challenges in
providing fast, reliable internet access to their
customers, due to the shared use of network
resources and the limited availability of
spectrum. Network Slicing may provide
options of how mobile operator networks can
better meet customer needs.
As the net neutrality debate has evolved,
policymakers have come to accept that
network management plays an important role
in service quality.
Industry position –
To meet the varying needs of consumers,
mobile network operators need the ability
to actively manage network traffic and the
flexibility to differentiate between different
types of traffic.
Regulation that affects network operators’
handling of mobile traffic is not required.
Any regulation that limits their flexibility to
manage network, service quality and
provide consumers with a satisfactory
experience is inherently
counterproductive.
Consumers should have the ability to
choose between competing service
providers and a variety of offers / tariffs to
best meet their individual needs.
Mobile network operators compete along
many dimensions, such as pricing of
service packages and devices, different
calling and data plans, innovative
applications and features, and network
quality and coverage. The high degree of
competition in the mobile market provides
ample incentives to ensure customers
enjoy the benefits.
Where there is flexibility, mobile network
operators are able to offer a bespoke,
managed service to providers of new
connected products, such as autonomous
cars, which could not exist without
constant, high-integrity connectivity.
Operators can also enter into commercial
arrangements with content and
application providers that want to attract
users by offering free access, that is,
zero-rating their content so mobile
subscribers are not ‘charged’ for the data
usage. These kinds of arrangements
enable product and service innovation,
deliver added value to consumers and
generate new revenue for network
operators, which face constant pressure
to enhance, extend and upgrade their
networks.
Quality of Service
The quality of a mobile data service is
characterised by a number of important
parameters, notably speed, packet loss,
delay and jitter. It is affected by factors such
as mobile signal strength, network load, and
user device and application design.
Industry Position
Competitive markets with minimal
regulatory intervention are best able to
49
deliver the quality of mobile service
customers expect. Regulation that sets a
minimum quality of service is
disproportionate and unnecessary. The
quality of service experienced by mobile
consumers is affected by many factors,
some of which are beyond the control of
operators, such as the device type,
application and propagation environment.
Defining specific quality targets is neither
proportionate nor practical.
Mobile networks are technically different
from fixed networks; they make use of
shared resources to a greater extent and
are more traffic-sensitive. Mobile network
operators need to deal with continually
changing traffic patterns and congestion,
within the limits imposed by finite network
capacity, where one user’s traffic can
have a significant effect on overall
network performance.
The commercial, operational and
technological environment in which mobile
services are offered is continuing to
develop. Mobile network operators must
have the freedom to manage and
prioritise traffic on their networks.
Regulation which rigidly defines a
particular service quality level is
unnecessary and is likely to impact the
development of these services.
Competitive markets with differentiated
commercial offers and information that
allows consumers to make an informed
choice deliver the best outcomes. If
regulatory authorities are concerned
about quality of service, they should
engage in dialogue with the industry to
find solutions that strike the right balance
on transparency of quality of service.
Cross-Border Data Transfers
The global digital economy depends on
cross-border data transfers to deliver crucial
social and economic benefits to individuals,
businesses and governments.
When data is allowed to flow freely across
national borders it enables organisations to
operate, to innovate and to access solutions
and support anywhere in the world. Enabling
cross-border data transfers can help
organisations adopt data-driven digital
transformation strategies that ultimately
benefit individuals and society. Policies that
inhibit the free flow of data through unjustified
restrictions or local data storage
requirements can have an adverse impact on
consumers, businesses and the economy in
general.
Cross-border transfers of personal data are
currently regulated by a number of
international, regional and national
instruments and laws intended to protect
individuals’ privacy, the local economy or
national security. While many of these
instruments and laws adopt common privacy
principles, they do not create an
interoperable regulatory framework that
reflects the realities, challenges and potential
of a globally connected world. Emerging
frameworks such as the APEC Cross-Border
Privacy Rules and the EU’s Binding
Corporate Rules allow organisations to
transfer personal data generally under certain
conditions. These frameworks contain
accountability mechanisms and are based on
internationally accepted data protection
principles.
However, their successful adoption is
undermined by the implementation by
governments of ‘data localisation’ (also
known as ‘data sovereignty’) rules that
impose local storage requirements or use of
local technology. Such localisation
requirements can be found in a variety of
sector- and subject-specific rules including for
financial service providers, the public sector
or professional confidentiality reasons and
are sometimes imposed by countries in the
belief that supervisory authorities can more
easily scrutinise data that is stored locally.
50
It is envisaged that network slices may be
utilised to offer services outside of the home
jurisdiction, potentially in a similar manner to
international roaming. Any transfer of data
across borders may need to consider
regulatory requirements.
Industry Position
Cross-border transfers of data play an
important role in innovation, competition
and economic and social development.
Governments can facilitate cross-border
data flows in a way that is consistent with
consumer privacy and local laws by
supporting industry best practices and
frameworks for the movement of data and
by working to make these frameworks
interoperable. Governments can also
ensure that these frameworks have strong
accountability mechanisms, and that the
authorities can play a role in
overseeing/monitoring their
implementation. Governments should only
impose measures that restrict cross-
border data flows if they are absolutely
necessary to achieve a legitimate public
policy objective. The application of these
measures should be proportionate and
not arbitrary or discriminatory against
foreign suppliers or services. mobile
network operators welcome frameworks
such as the APEC Cross-Border Privacy
Rules or the EU’s Binding Corporate
Rules, which allow accountable
organisations to transfer data globally,
provided they meet certain criteria. Such
mechanisms are based on commonly
recognised data privacy principles and
require organisations to adopt a
comprehensive approach towards data
privacy. This encourages more effective
protection for individuals than formalistic
administrative requirements while helping
to realise potential social and economic
benefits. Such frameworks should be
made interoperable across countries and
regions to the greatest extent possible in
order to seek convergence between
different approaches to privacy while
promoting appropriate standards of data
protection and to allow accountable
companies to build scalable and
consistent data privacy programmes.
Requirements for companies to use local
data storage or technology create
unnecessary duplication and cost for
companies and there is little evidence that
such policies produce tangible benefits for
local economies or improved privacy
protections for individuals.
To the extent that governments need to
scrutinise data for official purposes,
mobile network operators would
encourage them to achieve this through
existing lawful means and appropriate
intergovernmental mechanisms that do
not restrict the flow of data.
Illegal Content
Today, mobile networks not only offer
traditional voice and messaging services, but
also provide access to virtually all forms of
digital content via the internet. In this respect,
mobile network operators offer the same
service as any other internet service provider
(ISP). This means mobile networks are
inevitably used, by some, to access illegal
content, ranging from pirated material that
infringes intellectual property rights (IPR) to
racist content or child sexual abuse material
(child pornography).
Laws regarding illegal content vary
considerably. Some content, such as child
sexual abuse material, is considered illegal
around the world, while other content, such
as dialogue that calls for political reform, is
illegal in some countries while being
protected by ‘freedom of speech’ rights in
others.
Communications service providers, including
mobile network operators and ISPs, are not
usually liable for illegal content on their
networks and services, provided they are not
aware of its presence and follow certain rules
51
e.g., ‘notice and take-down’ processes to
remove or disable access to the illegal
content as soon as they are notified of its
existence by the appropriate legal authority.
Mobile network operators are typically alerted
to illegal content by national hotline
organisations or law-enforcement agencies.
When content is reported, operators follow
procedures according to the relevant data
protection, privacy and disclosure legislation.
In the case of child sexual abuse content,
mobile network operators use terms and
conditions, notice and take-down processes
and reporting mechanisms to keep their
services free of this content.
The issues in relation to illegal content,
liability, regulatory and self-regulatory
obligations and commitments will need to be
addressed with the introduction of Network
Slicing. These may also vary depending on
how slices are implemented.
Industry Position
The mobile industry is committed to
working with law enforcement agencies
and appropriate authorities, and to having
robust processes in place that enable the
swift removal or disabling of confirmed
instances of illegal content hosted on their
services. ISPs, including mobile network
operators, are not qualified to decide what
is and is not illegal content, the scope of
which is wide and varies between
countries. As such, they should not be
expected to monitor and judge third-party
material, whether it is hosted on, or
accessed through, their own network.
National governments decide what
constitutes illegal content in their country;
they should be open and transparent
about which content is illegal before
handing enforcement responsibility to
hotlines, law-enforcement agencies and
industry.
The mobile industry condemns the misuse
of its services for sharing child sexual
abuse content. The GSMA’s Mobile
Alliance Against Child Sexual Abuse
Content provides leadership in this area
and works proactively to combat the
misuse of mobile networks and services
by criminals seeking to access or share
child sexual abuse content.
Regarding copyright infringement and
piracy, the mobile industry recognises the
importance of proper compensation for
rights holders and prevention of
unauthorised distribution.
Summary
At this stage there do not seem to be new
regulatory challenges as a result of Network
Slicing. However, the challenge(s) may be
on how best to meet existing regulatory
obligations as result of Network Slicing.
Implementation of Network Slicing may also
need to consider a range of other regulatory
issues from, but not limited to - access to
emergency calls, support of law enforcement
and government access to data and the
related legislative frameworks that enables
this, to consumer protection and competition
policy.
52
7 Next Steps
53
Next steps
To date within GSMA we have looked at the
Network Slicing topic as a new capability that
will be delivered with the new 5G core
network. We still have considerable work to
do, including; definition of slice types and the
parameter sets that may exist within different
slices, validate 3GPP defined Slices,
coordinate industry Slicing activities, explore
Business Models, market sizes, market
opportunities and value chains, etc. This will
form basis for the deliverables in the next
phase. We welcome input from Operators,
Vendors and customers into this technical
paper.
Also, many operators are not waiting for 5G
Core (3GPP Release16) to explore the
concept of delivering differentiated
capabilities to different customer sets.
Operators have different physical and
increasingly virtual networks and network
capabilities today, including; 2G, 3G, 4G, NB-
IoT, Emergency Services LTE, etc.
Virtualisation of network elements facilitates
the automation of network orchestration,
configuration and deployment. Current
capabilities like Edge Cloud allow us to start
to explore lower latency opportunities, and in
the near future we will begin to add 5G New
Radio to serve new and existing customers.
We will also use 5G NR to deliver against
MNO requirements like Fixed Wireless
Access substitution. Operators are beginning
to differentiate between Consumer focused
networks and Business or Vertical focused
networks.
While 5G core and the real slicing capability
allow for significantly more efficient
deployment of sliced capability, orchestration
and configuration mechanisms, operators are
willing to explore opportunities for service
differentiation and optimisation offered by
pre-5G solutions.
Consequently, another deliverable in the next
phase will be a paper focusing on the
emerging (or bottom up) trend of using
Multiple Dedicated Core Networks (MDCN) to
deliver differentiated services and solutions
utilising existing EPS system capabilities. We
anticipate this paper to be available at the
end of Q318. We welcome input from
Operators, Vendors and customers.
Finally, it is important to continue the
interactions with the vertical industries in
order to better understand their use cases
and requirements.
54
Annex A Document Management
A.1 Document History
Version Date Brief Description of Change Approval
Authority
Editor /
Company
1.0 New Future Network Programme
document
Future Network
Programme
Network Slicing
Taskforce
Future Network
Programme
Network
Slicing
Taskforce
A.2 Other Information
Type Description
Document Owner GSMA Future Networks Programme
Editor / Company Michele Zarri / GSMA
Kelvin Qin/ GSMA
Reviewed & Approved by Future Network Programme Network Slicing Taskforce
It is our intention to provide a quality product for your use. If you find any errors or omissions,
please contact us with your comments. You may notify us at [email protected]
Your comments or suggestions & questions are always welcome.
